ABSTRACT
Highways are supposed to be of paramount importance infrastructure for any nation. But, potholed laden roads, poor weather conditions or man-made errors results in road crashes, and the ones due to fog cause greater severity to life. To address this problem, we have proposed a model in which the images captured by cameras embedded in the vehicles are processed at real time to restore some visibility. The proposed solution is to display improvised dehazed images of the road scenario to help drivers anticipate potential collisions based on wavelet-based image processing. The future scope of this study is applicable in Advanced Driver Assistance Systems in fog dominant areas. To evaluate the model’s performance fairly and objectively, the model is trained with two image datasets with different visibility ranges. The comparative study and quantitative evaluation exhibit the proposed method can obtain very good defogging image with relatively fast speed.
